ORIGINAL PAPER: NANO- AND MACROPOROUS MATERIALS (AEROGELS, XEROGELS, CRYOGELS, ETC.)

Flexible acrylate-grafted silica aerogels for insulation purposes:comparison of reinforcement strategies

João P. Vareda1 • Telma Matias1 • Ana C. Fonseca2 • Luisa Durães1

Received: 4 March 2016 / Accepted: 26 June 2016

� Springer Science+Business Media New York 2016

Abstract Vinyltrimethoxysilane (VTMS), methyltrime-

thoxysilane (MTMS) and tetramethylorthosilicate (TMOS)

were mixed in a one-step basic catalyzed sol–gel chemistry, to

produce flexible and good thermal insulator silica-based aero-

gels. Moreover, mechanical reinforcement of the aerogels with

a polymer phase was accomplished via free radical polymer-

ization using either the macromer poly(ethylene glycol)

diacrylate (PEG-DA) or the monomer 2-(dimethylamino)ethyl

methacrylate (DMAEMA), the latter case resulting in the

grafting of PDMAEMA. Both the influence of using a mono-

mer or a macromer and of applying two different polymer

grafting approaches—one-pot synthesis or gel soaking—on the

aerogels’ final properties were analyzed. It was concluded that

gel soaking strongly limits the diffusion of the organic moieties,

creating heterogeneous materials. This approach led to denser

and less hydrophobic aerogels with worse mechanical proper-

ties. The incorporation of low amounts of polymer with one-pot

synthesis led to an improvement in the aerogels properties,

making them less dense, more flexible and better insulators.

The one-pot method also allowed to obtain aerogels with a very

well-defined microstructure, due to the porogen role of the

polymer, hence generating homogenous materials. It was found

that the major differences in the aerogels properties occurred

with different grafting approaches and not with different

polymer phases, and the molecular weight of the organic

moieties was not determinant for the materials homogeneity, in

the studied cases. The obtained bulk densities ranged from 122

to 181 kg m-3, the thermal conductivity from 0.050 to

0.072 W m-1 K-1 and the modulus achieved 327 kPa. The

addition of DMAEMA generated the best results when using

the one-pot approach, providing the lightest and uncommonly

flexible thermal insulator aerogels (modulus of *70 kPa).

Graphical Abstract

Keywords Silica aerogel � Grafting � One-pot � Gelsoaking � Acrylate � Thermal insulation

1 Introduction

Silica aerogels are highly porous, lightweight and good

thermal insulator materials obtained by supercritical drying

techniques [1–5]. Moreover, these materials can exhibit

other excellent and unique properties, which can be

Electronic supplementary material The online version of thisarticle (doi:10.1007/s10971-016-4137-6) contains supplementarymaterial, which is available to authorized users.

& Luisa Durãesluisa@eq.uc.pt

1 CIEPQPF, Department of Chemical Engineering, University

of Coimbra, Rua Sı́lvio Lima, 3030-790 Coimbra, Portugal

2 CEMUC, Department of Chemical Engineering, University

of Coimbra, 3030-790 Coimbra, Portugal

123

J Sol-Gel Sci Technol

DOI 10.1007/s10971-016-4137-6

http://dx.doi.org/10.1007/s10971-016-4137-6http://crossmark.crossref.org/dialog/?doi=10.1007/s10971-016-4137-6&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1007/s10971-016-4137-6&domain=pdf

tailored using sol–gel technology. For example,

methyltrimethoxysilane (MTMS)- and methyltriethoxysi-

lane (MTES)-derived aerogels show uncommon high

flexibility and hydrophobicity [3, 4, 6]. Thus, silica aero-

gels have been proposed for building insulation (for both

thermal and acoustic insulation) [5, 6], thermal insulators

for space vehicles and extra-vehicular activity [7], catalyst

supports [6], supercapacitors [2] and Cherenkov radiation

detectors [1], among others. Nonetheless, silica aerogels

usually have weak mechanical properties that limit their

handling/processing and use in load bearing applications.

Several strategies have been applied to overcome the

fragile nature of silica aerogels [8]. The inclusion of flex-

ible bis-silanes into the silica backbone, via co-gelation,

has shown to reduce the Young’s modulus and bulk density

of the resulting aerogels when using optimized conditions

[7–10]. The addition of these precursors creates a more

open structure that also leads to better thermal insulation

characteristics [11].

A different approach to the problem is the reinforcement

of the silica structure by cross-linking with polymers,

obtaining hybrid materials [8, 9, 12]. The polymeric chains

grow on the modified surface of secondary particles, cre-

ating a coating and increasing the connection between them

[11]. This strategy strengthens the pearl necklace structure

of the mesoporous silica making it sturdier. However, this

also increases the aerogel’s bulk density which leads to an

increased thermal conductivity [8]. Regardless of being

two different approaches Nguyen et al. [13], Meador et al.

[14], Maleki et al. [11] and Randall et al. [7] combined the

two. The latter studied the properties obtained with dif-

ferent silica backbones for the same cross-linker and the

effect of different cross-linkers on the properties of the

same silica backbone. Randall et al. [7] concluded that

larger alkyl chains in the bis-silane improved some desired

properties (low bulk density, high modulus and recovery

from compression), with the hexyl link being considered an

optimum chain length. This combined approach allows, in

theory, to improve the aerogels mechanical properties even

further, and the open structure obtained in the silica

backbone with the co-gelation of bis-silanes compensates,

to some extension, the increasing density due to cross-

linking.

Reinforcement of silica aerogels via polymer cross-

linking has been widely investigated, and a variety of

different cross-linkers like epoxide [7, 14, 15], poly

(acrylates) [7, 8, 12, 16], polyureas [7, 8], isocyanate

[7, 8, 13] and polystyrene [7–9, 12, 16] have been used for

reinforcement. To create a covalent bond between the two

phases, functionalization of the silica backbone is neces-

sary and can be performed by co-gelation with a silane

containing an organic reactive functional group, like vinyl

or amine [8, 9, 14]. Liquid-phase polymer reinforcement,

the most frequent strategy, can be achieved in two different

ways. The first consists in soaking the aged alcogel in a

monomer solution prepared with a post gelation solvent.

During the soaking period, the monomer molecules diffuse

into the interior of the wet gel. This method is time-con-

suming and is limited by monomer diffusion through the

highly complex pore system of the silica aerogel [8, 9, 14].

The second method, also called one-pot synthesis, is less

time-consuming and more straightforward, with the

organic moieties being added to the sol along with the

silica precursors. With this strategy, the monomer mole-

cules are entrapped inside the porous structure. Afterward,

for both cases, the reaction between the monomer and

functionalized silica is promoted by a heat stimulus and the

resulting hybrid wet gel is dried [8, 11, 15].

In this study, we test the polymer reinforcement of silica-

based aerogels obtained from the co-gelation of

vinyltrimethoxysilane (VTMS), trimethoxymethylsilane

(MTMS) and tetramethylorthosilicate (TMOS). With the

used mixture of silane alkoxides, we intend to synthesize

flexible and hydrophobic aerogels, in a one-step basic

catalysis, which also include organic moieties (vinyl groups)

that can react with the monomers via free radical polymer-

ization. Poly(ethylene glycol) diacrylate (PEG-DA) and

2-(dimethylamino)ethyl methacrylate (DMAEMA) were

studied as macromer and monomer for polymer reinforce-

ment. The different molecular weight will influence mono-

mer diffusion in gel soaking methodology and the porogen

effect in one-pot synthesis. Since PEG-DA is highly

hydrophilic, it is important to maintain it in low amount to

not compromise the hydrophobicity of the final material.

Monoliths obtained with both gel soaking and one-pot

synthesis are compared in order to determine the most effi-

cient method for the manufacturing of monoliths for insu-

lation-related applications. Key properties of the aerogels

for these applications, such as bulk density (porosity),

thermal conductivity, flexibility and hydrophobicity, were

evaluated.

2 Experimental

2.1 Materials

Vinyltrimethoxysilane (VTMS, 98 %, H2C=CHSi(OCH3)3),

methyltrimethoxysilane (MTMS, 98 %, CH3Si(OCH3)3),

tetramethylorthosilicate (TMOS, C98 %, Si(OCH3)4),

poly(ethyleneglycol diacrylate) (PEG-DA, Mn *575, H2C=CHCO(OCH2CH2)nOCOCH=CH2), 2-(dimethylamino)

ethyl methacrylate) (DMAEMA, CH2=C(CH3)COO

CH2CH2N(CH3)2), methanol (C99.8 %, CH3OH), tetrahy-

drofuran (THF, C99.9 %, C4H8O), ammonium hydroxide

(25 % wt, NH4OH) and 2,20-azobis(2-methylpropionitrile)

J Sol-Gel Sci Technol

123

(AIBN, &98 %, (CH3)2C(CN)N=NC(CH3)2CN) were pur-chased from Sigma-Aldrich. Only DMAEMA was purified

prior to utilization by passing it over a silica–alumina column.

2.2 Synthesis of the silica backbone

Preliminary tests were conducted in order to optimize the

silica precursors’ mixture. Selection criteria included

complete gelation of the mixture, bulk density and mono-

lithicity [17].

A solution consisting of VTMS (40 mol%), MTMS

(40 mol%) and TMOS (20 mol%) in methanol was stirred

at 0 �C in an ice/ethanol bath. A 1 M aqueous solution ofammonium hydroxide was prepared separately and added

to the first one. Typically, for 100 mL of gel, 1 mL of

ammonium hydroxide solution (25 % in water) was added

to the precursors’ mixture, keeping a total silicon concen-

tration of 1.82 M and a water/silicon molar ratio of 7.8.

The resulting mixture was homogenized for a few seconds,

then, poured in cylindrical poly(propylene) molds and left

for aging at 27 �C for 48 h.

2.3 Polymer reinforcement via gel soaking

After the aging period, the silica alcogels were unmolded

and dipped in a new solution made of a post-gelation sol-

vent (THF), the monomer/macromer to be grafted and

AIBN, with 50 % (w/w) monomer/macromer to THF. The

amount of AIBN added was 3 % of monomer/macromer

mass. For DMAEMA, this solution was degassed for

30 min with nitrogen. After being soaked for 72 h at room

temperature, the wet gels were placed in a cylindrical

reaction flask, covered with fresh THF and then refluxed

for 24 h for polymerization (Fig. 1).

The amount of monomer added to the system was based

on its molecular weight and prior results [17]. Table 1

features the monomer concentrations tested in the gel

soaking procedure. PEG-DA already has a developed car-

bon chain, and for reinforcement purposes, it is not crucial

that it undergoes further growing. Moreover, each mole-

cule of this polymer has two reactive ends that can link to

vinyl groups, so the PEG-DA/VTMS molar ratio was kept

low, although with a slight excess to increase the concen-

tration difference for a more efficient diffusion. On the

other hand, in the case of DMAEMA, the monomer/VTMS

ratio was increased in order to favor polymer chain growth.

2.4 Preparation of reinforced gels via one-pot

synthesis

An aqueous solution of ammonium hydroxide 1 M was

combined with methanol, the monomer/macromer and

AIBN (3 % of the monomer/macromer mass). This was

added, under stirring, to a solution containing methanol and

the silica precursors. After stirring, the mixture was poured

into cylindrical poly(propylene) molds at room temperature

for gelation and then left for aging at 27 �C for 48 h. Afteraging, the polymerization of the monomers in the gel

proceeded applying THF reflux, as mentioned in Sect. 2.3

(Fig. 2).

In order to optimize the monoliths properties, some

monomer concentrations were tested as there were no

previous results [17]. Table 1 presents the tested monomer/

VTMS molar ratios. Concentrations were varied, starting

on a small monomer/VTMS molar ratio and then increased

until gelation was inhibited or the resulting monolith

became very rigid.

The reference sample, which is an unreinforced aerogel

with the same modified silica backbone of the reinforced

Fig. 1 Schematicrepresentation of the

experimental procedure used for

grafting via gel soaking

J Sol-Gel Sci Technol

123

aerogels, is labeled as A_P0. The reinforced aerogels

labeling uses the following scheme: A stands for aerogel;

PX identifies the type of polymer grafted in the silica

network (X = P, aerogel reinforced with PEG-DA;

X = D, aerogel reinforced with PDMAEMA); S means

that gel soaking was used as grafting method, while O

corresponds to one-pot synthesis; and the last term defines

the molar ratio (R) between monomer and VTMS (for

example, R2 means that for each vinyl group of the silica

backbone there are two monomer molecules in the system).

2.5 Gels drying

Gels were placed in a supercritical extraction stainless steel

chamber and dried via continuous supercritical fluids

extraction. The gels were firstly washed with a flow of hot

methanol at high pressure, and then, they were dried by

extracting the methanol with supercritical CO2 at 50 �Cand 150 bar [18].

2.6 Aerogels characterization

Key properties of the aerogels were determined as follows:

(1) bulk density—by weight–volume measurements of four

cubic pieces of each replica; (2) porosity—calculated from

bulk and skeletal densities, using Eq. (1), being the skeletal

density accessed with helium pycnometry (Accupyc 1330,

Micrometrics) carried out on previously ground samples;

(3) hydrophobicity—determined via contact angle mea-

surement (OCA 20, Dataphysics), using the sessile drop

technique with high purity water as test liquid; (4) thermal

conductivity—measurements were taken using the tran-

sient plane source method (TPS2500 Hot Disk Thermal

Constants Analyzer, Hot Disk) and placing the sensor

between the flat side of two cylindrical pieces of material

with similar dimensions, at 20 �C.

Porosity % ¼ 100 � 1 � qbqs

� �ð1Þ

Table 1 Monomer/VTMS molar ratios (R) tested for gel soaking and one-pot approaches

Method of monomer addition PEG-DA/VTMS (mol/mol) Samples label DMAEMA/VTMS (mol/mol) Samples label

Gel soaking 1.6 A_PP_S_R1.6 10 A_PD_S_R10

One-pot 0.5 A_PP_O_R0.5 0.5 A_PD_O_R0.5

1 A_PP_O_R1 1 A_PD_O_R1

2 A_PD_O_R2

5 A_PD_O_R5

Fig. 2 Schematicrepresentation of the

experimental procedure used for

grafting via one-pot

J Sol-Gel Sci Technol

123

Chemical structure of the aerogel monoliths was accessed

by FTIR spectroscopy (FT/IR 4200, Jasco), using the KBr

mode, in the range of 4000–400 cm-1, with a resolution of

4 cm-1 and 64 scans.

Thermal characterization was performed by thermo-

gravimetric analysis—TGA (SDTQ600, TA Instruments),

using powdered samples, until a temperature of 1200 �C, at10 �C min-1 in a nitrogen atmosphere.

Mechanical behavior of the monoliths was accessed apply-

ing a uniaxial compression test (Shimadzu AG-IS 100KN,

control of the compression velocity and a AND EK-1200i bal-

ance) on cylindrical samples, with a speed of 1.3 mm/min.

Microstructure was observed and chemically characterized

with field emission scanning electron microscope (Compact/

VPCompact FESEM, Zeiss Merlin) and energy-dispersive

X-ray spectroscopy (EDS) (XMaxN, Oxford), respectively.

3 Results and discussion

Most promising aerogels, both non-reinforced and rein-

forced with PEG-DA and PDMAEMA, are shown in

Fig. 3. Gelation for both non-reinforced and grafted gels

was rather quick, with gelation occurring within 5 min

after the mixing of the basic catalyst. Three replicas of each

sample were synthesized and characterized, except for

sample A_PP_S_R1.6.

Non-cross-linked silica aerogels exhibited a shining,

plastic look on its outer surface, and, although being

relatively rigid, they were still compressible. Moreover,

they show negligible particle shedding. Aerogels grafted

using the gel-soaking technique have a very different

aspect. The monoliths are more rigid in general, and when

grafted with PEG-DA, they show heterogeneity, with

rigid and soft parts, and could be easily broken. The

shining look was not observed in this case, but the

aerogels remained quite clean in terms of particles

release. One-pot synthesis originated the most flexible

monoliths. These show a dull-looking surface, and they

always released powder.

It was observed that for a higher amount of polymer the

aerogel’s integrity improved, although there is a concen-

tration for which the aerogel became extremely rigid.

3.1 Key properties

Table 2 presents the key properties of the most promising

monoliths. The mechanical properties presented in this

table will be discussed only in Sect. 3.4.

It is evident that grafting via gel soaking, comparing

with non-reinforced aerogels, generally creates denser

Fig. 3 Photographs of some ofthe monoliths. a A_P0 (non-reinforced sample);

b A_PP_S_R1.6;c A_PP_O_R0.5;d A_PD_S_R10;e A_PD_O_R2

J Sol-Gel Sci Technol

123

aerogels with increased thermal conductivity and decreased

porosity. Porosity is related to the bulk density, being

inversely proportional properties and is calculated with

both bulk and skeletal densities by Eq. (1). Thus, it is

influenced by the uncertainties of these two measurements

and by the amount of closed pores in the structure. In this

way, only averaged nominal values are presented.

Shrinkage, also presented in Table 2, was estimated

using the theoretical and bulk densities of samples. For the

theoretical density calculation, it was assumed that the gel

skeleton mass remains constant during drying, being the

theoretical mass obtained considering complete hydrolysis

and condensation of silica precursors and the polymer mass

percentage measured by TGA analysis. The volume of

solution, hence the gel volume, was kept at 40 mL in all

syntheses.

Shrinkage of the gels is fairly similar in all samples, as

expected considering their monolithic appearance from

Fig. 3. However, it seems that the addition of a polymer

phase tends to decrease the gel’s shrinkage during super-

critical drying, which is more noticeable in the samples with

PEG-DA. In fact, this polymer tends to cross-link, giving a

more cohesive structure that hinders shrinkage. When in one-

pot layout, this cross-linking is not so extensive due to the

presence of the silica structure, thus giving higher shrinkage

values than for the gel-soaking approach.

Thermal conductivity of aerogels, at low temperature, is

mainly dependent on solid heat conduction through the

silica particles and heat conduction through the gaseous

phase [18, 19]. The gaseous phase can transfer heat through

conduction and convection, being the latest inhibited/re-

duced due to the small pore dimension and complex porous

network of aerogels [19]. Thus, aerogels thermal conduc-

tivity is usually the sum of the solid heat conduction and

gas heat conduction [18, 19], and the latter is the limiting

step. Therefore, denser aerogels tend to be worse insula-

tors, being our results in agreement with this trend

(Table 2).

The addition of the organic moieties in the sol phase

(one-pot strategy) created a different aerogel structure.

Results suggest a tendency for weightless and better insu-

lator monoliths in this case (Table 2), characteristics

originated by the porogen effect of the polymer during

gelation, which controls particle growth and decreases the

pore size (as it will be seen later in SEM analysis),

decreasing the influence of the gaseous convection but also

reducing the bulk density.

Hydrophobic monoliths are stable when exposed to

moisture. Methyl and vinyl groups derived from MTMS

and VTMS, respectively, turned the silica backbone

hydrophobic, as shown by the high contact angle values of

the synthesized aerogels (112–150�, Table 2). However,not all obtained aerogels are equally hydrophobic. Gel-

soaking-grafted aerogels show a decrease in contact angle

(112� and 122�), when compared to non-reinforced aero-gels (A2_P0, 144�), and thus, those were considered asbeing near the transition between hydrophilic and

hydrophobic materials. In this type of aerogels, for PEG-

DA reinforcement, the water drop was often absorbed by

the test material which caused difficulties in the analyses;

this was certainly due to the hydrophilic nature of PEG-DA

and its heterogeneous distribution on the aerogel surface.

Grafting via one-pot retains the hydrophobic character-

istics of the aerogels, probably due to a more organized

porous structure provided by the porogen effect of the

organic moieties, which leads to well-dispersed

hydrophobic functional groups.

For both polymers, one-pot synthesis is the only grafting

method that improves the discussed key properties, creat-

ing the most adequate materials, from this point of view.

Moreover, it is PDMAEMA that most improves these

properties, being the sample A_PD_O_R2 the best one in

this context.

The aerogels prepared in this study are lighter than the

ones obtained with epoxy cross-linking that feature bulk

densities varying from 198 to 847 kg m-3 [15]. The

polystyrene-grafted materials obtained by Nguyen et al. [9]

also feature a wide range of bulk densities, going up to

370 kg m-3. The lightest materials obtained by these

authors have similar densities to the ones achieved in this

Table 2 Key properties of selected aerogels

Sample Bulk density

(kg m-3)

Porosity

(%)

Shrinkage

(%)(a)Contact

angle (�)Thermal

conductivity

(W m-1 K-1)

Young’s

modulus

(kPa)

Compressive

strength (kPa)

Elongation at

break (%)

A_P0 148.8 ± 5.6 89 13.7 144 ± 12 0.063 ± 0.004 927.5 71.3 16.37

A_PP_S_R1.6 147.5 88 5.4 112 0.072 627.1 47.0 15.50

A_PP_O_R0.5 132.7 ± 7.6 91 10.6 151 ± 6 0.061 ± 0.017 97.6 19.3 32.76

A_PD_S_R10 181.5 ± 50.5 86 12.1 122 ± 56 0.067 ± 0.022 848.4 366.6 47.82

A_PD_O_R2 122.6 ± 6.8 91 13.4 150 ± 9 0.050 ± 0.007 68.6 19.5 30.17

a Shrinkage was estimated by the bulk and theoretical densities using the formula: 1 � qtqb� �

� 100

J Sol-Gel Sci Technol

123

publication via the one-pot method. Maleki et al. [16]

grafted aerogels with poly(butyl acrylate) and polystyrene,

obtaining monoliths of similar density.

In spite of having interesting insulation properties, the

aerogels with polymer reinforcement presented in this

paper are not yet as good insulators as MTMS-derived

aerogels and aerogel-like materials [18, 20], as expected

due to the much lower bulk density of the latter.

3.2 Chemical structure

3.2.1 FTIR

Figure 4 shows the typical FTIR spectra for the synthesized

aerogels, from which the characteristic functional groups

were identified considering the data in the open literature

[21]. The same chemical system, although with variation in

the relative proportion of its components (different sam-

ples) or in the grafting method, generated similar spectra.

Thus, a representative spectrum was selected to show the

typical functional groups of the aerogels for each specific

chemical system.

For all gels, an organically modified silica network is

formed, whose chemical composition is in agreement with

an aerogel derived from MTMS, VTMS and TMOS [21].

There is a predominance of siloxane (Si–O–Si) bonds,

being also observed the methyl and vinyl groups incorpo-

rated therein as well as the hydroxyl groups from network

ends. The characteristic vibrations of the siloxane bonds

are detected at approximately 438 cm-1 (bending),

540 cm-1 (SiO2 defects), 771 cm-1 and between 1039 and

1135 cm-1 (stretching). The vibrations of the methyl

groups are also visible in all aerogels as this functional

group is not reactive. The vibrations of the vinyl groups

can be found at 540, 920, 1002, 1383, 1412, 1604, 1940,

3027, 3065 cm-1, while peaks representing vibrations

associated with the methyl groups can be found at 835,

1276, 1412, 1477, 2839, 2875 and 2982 cm-1. Peaks rep-

resenting the hydroxyl group are found at 1630 and

3447 cm-1.

For grafted aerogels, the spectra show bands that cor-

respond to the characteristic functional groups of each

polymer (C–N stretching for PDMAEMA, C–O–C

stretching and C=O stretching for both polymers) that are

sometimes overlapped with other peaks, but also show that

the vinyl groups still exist (peak at 1600 cm-1). We can

conclude from this observation that the remaining vinyl

groups are either from remaining active sites on the silica

backbone (unreacted VTMS) or from unreacted monomer

molecules. Thus, not all vinyl groups were used to create

covalent bonds between the two phases. This can be due to

the highly complex pore system of the aerogel, which

hinders the diffusion of the organic moieties to reach the

vinyl groups on the silica backbone, and also due to steric

hindrance caused by the methyl groups.

3.3 Thermal characterization

TGA analysis was conducted for selected aerogels and

grafting polymers (Fig. 5). Monomers were polymerized

using the same conditions applied to the gels heat treat-

ment, and the neat polymer was also subjected to the

Fig. 4 Typical FTIR spectra forthe synthesized aerogels (t—stretching vibration; d—bending vibration; s—

symmetric; as—asymmetric;

c—out of plane bendingvibration; in particular, the

shown spectra are from samples

A_P0, A_PP_O_R0.5 and

A_PD_S_R10)

J Sol-Gel Sci Technol

123

thermal degradation analysis. Tests were ran with a sample

representative of the entire aerogel (mixing the core

material with that of the outer surface), but for gel-soaking-

grafted aerogels, a sample representative of the core (called

interior) and one from the outer surface (called exterior)

were analyzed separately. Table S1 shows the stages of

mass loss and respective temperature ranges. The average

amount of polymer mass in each grafted aerogel was

obtained running three independent tests.

Figure 5a shows that PEG-DA-grafted aerogels

(A_PP_S_R1.6 and A_PP_O_R0.5) start to degrade at

higher temperatures when compared to the non-reinforced

aerogel (A_P0) or the neat polymer itself. The covalent

bonds between carbon and silicon atoms are very stable;

hence, the hybrid aerogels show increased thermal stabil-

ity. PEG-DA-grafted aerogels undergo significant mass

loss between 400 and 600 �C (Fig. 5a, Table S1), attributedto the polymer-phase degradation. The differences between

residual masses (in relation to the non-reinforced aerogel)

and the shape of the curve itself reveal the incorporation of

the polymer phase in reinforced aerogels.

Representative curves of outer surface and core of the

hybrid aerogels prepared by gel soaking allow under-

standing the efficiency of monomer diffusion through the

gel and the heterogeneity of these materials. These two

curves feature a significant difference in the amount of

polymer, meaning that this methodology creates hetero-

geneous materials due to difficult monomer diffusion, as

reported before [15]. It is reasonable to assume that poly-

mer concentration in the aerogel varies in the radial

direction, decreasing toward the center of the cylindrical

samples. Comparing the hybrid aerogels obtained by one-

pot and gel soaking strategies, it is notorious that one-pot

synthesis succeeds in incorporating more PEG-DA in the

aerogel (8.90 ± 1.97 % for A_PP_O_R0.5 and

7.98 ± 0.69 % for A_PP_S_R1.6), considering that the

amount of monomer added in this case was much smaller

than for gel soaking.

PDMAEMA-grafted aerogels (A_PD_S_R10 and

A_PD_O_R2) feature a similar behavior. These aerogels

are also more stable than the individual phases (Fig. 5b). A

significant weight loss is visible between 300 and 550 �C(Fig. 5b, Table S1), corresponding to polymer-phase

degradation. Outer surface and core curves reveal that

diffusional limitation and heterogeneities are still present

when using a monomer. We can conclude that although

less significant in PDMAEMA-grafted aerogels, the diffu-

sional limitations also happen in gel-soaking-grafted

aerogels, even with DMAEMA’s smaller molecular

weight. In these aerogels, gel soaking was the methodology

that allowed to incorporate more polymer into the monolith

(2.95 ± 1.63 % for A_PD_O_R2 one-pot and

19.59 ± 5.26 % for A_PD_S_R10) and this result is now

clearly dependent on the amount of monomer added to the

system. However, one-pot synthesis creates less hetero-

geneities and allows for polymer incorporation even with

small amounts of organic moieties added.

3.4 Mechanical behavior

Not all uniaxial compression tests ended with complete

breaking of the samples. Instead, the material started to

peal, with fractures being created at the exterior surface,

confirming the heterogeneous nature of these materials.

This behavior was expected because the outer surface of

the aerogels is denser, with a less open structure, due to

more extensive gelation next to the solid surface of the

container. Moreover, in the soaked aerogels, the amount of

polymer in the outer surface is higher than the amount on

Fig. 5 TGA curves of the synthesized aerogels; a PEG-DA-graftedaerogels versus non-reinforced aerogel; b PDMAEMA-grafted aero-gels versus non-reinforced aerogel. Samples representative of the

aerogels core are labeled as ‘‘int’’ (=interior) and of the outer surface

as ‘‘ext’’ (=exterior)

J Sol-Gel Sci Technol

123

its interior, which increases the rigidity of this surface.

Figure 6 presents the stress–strain diagrams for the selected

samples, and Table 2 summarizes the corresponding

mechanical properties.

Table 2 shows that the most rigid aerogel, which has the

higher value of Young’s modulus, is the non-reinforced

one (928 kPa). This behavior was expected because the

polymer phase features covalent bonds that are much more

elastic than the siloxane bonds, helping the flexibility of the

structure. Although not all elastic modulus values were

significantly reduced, in general polymer reinforcement

decreased the materials rigidity, creating more flexible

aerogels (modulus B848 kPa for reinforced samples). This

behavior is more noticeable on aerogels grafted via one-pot

(69 kPa with DMAEMA, 98 kPa with PEG-DA).

Polymer reinforcement should have also increased the

compressive strength of the materials, but this fact was not

observed. Because of the peeling behavior of monoliths

during the test, these measurements do not give the

mechanical strength of the aerogel’s core part that is cer-

tainly much higher. One-pot-grafted aerogels are the ones

that present the smaller compressive strength, regardless

the polymer used (*20 kPa). The presence of the organicmoieties during gelation creates more homogeneous

monoliths but makes a more fragile silica structure, due to

less extent of condensation. The significant mass loss after

1000 �C that these materials suffer (Fig. 5, Table S1),comparing to the others, is consistent with defects of the

silica network.

Materials that are less rigid can deform more until break,

and thus, polymer reinforcement should also increase the

materials elongation at break. Table 2 shows that grafted

aerogels have higher elongation at break (up to 48 %) than

the non-reinforced aerogel (16 %), with exception of

sample A_PP_S_R1.6 (16 %). Since the one-pot-grafted

aerogels are the most flexible materials, they present, in

general, high values of elongation at break (*30 %).Sample A_PD_S_R10 shows a particular behavior, as it

stood until a very high elongation at break (48 %),

although this sample was the stiffest of the reinforced ones

(modulus of 848 kPa). The presence of a high amount of

polymer (see Sect. 3.3) is the reason for this behavior.

Therefore, this confirms that the polymer phase can

increase significantly the compression strength (in this case

by more than fivefold), while maintaining the flexibility of

the materials at the same level, but has a deleterious effect

in the bulk density and thermal conductivity (Table 2).

Globally, grafting the aerogels with a polymeric phase

improves some of their mechanical properties, but this

depends on several factors, like the type and amount of

polymer, its distribution in the aerogel and, not less

important, the interference that it causes in the silica

polycondensation process during gelation.

Because of the ether functional groups, PEG-DA should

have created the most flexible aerogels. However, PEG-

DA-grafted aerogels feature a smaller Young’s modulus

only with gel soaking (627 vs. 848 kPa for the PDAEMA-

grafted sample), but in one-pot have a slightly higher

elongation at break, suggesting good flexibility (33 vs.

30 % for the PDAEMA-grafted sample).

Many authors [7, 9, 13–15] reported grafted silica

aerogels that have a modulus in the tens of MPa. In this

study, we developed aerogels whose modulus is in the tens

of kPa, creating aerogels that are much more flexible.

However, these are not yet as flexible as pure MTMS [18]

aerogels or as the grafted aerogels reported by Maleki et al.

[11], but the latest have bis-silanes in their composition,

creating a more open structure with more elastic bonds

within the backbone.

3.5 Microstructure

SEM images (Fig. 7) reveal spheroid secondary particles

with similar size, confirming the existence of the pearl

necklace structure (silica network) in all aerogels. In some

images, deposits from fragmentation of the particles or

from the gold coating can be noticed.

Non-reinforced and gel-soaking-grafted aerogels show a

similar pearl necklace structure: particles have a sphere-

like shape with similar sizes among them. Moreover, the

average particle size seems to be 5 lm for all these aero-gels. In spite of TGA showing the existence of polymer in

the core of the gel-soaking-grafted aerogels, SEM images

show relatively well-defined particles and necks. Thus, the

polymer is not noticeable at this level amount, being

completely mingled within the structure.

Fig. 6 Stress–strain diagrams of the selected aerogels

J Sol-Gel Sci Technol

123

As the TGA revealed significant differences between the

core and the outer surface of monoliths grafted through gel

soaking, these two parts were also observed by SEM sep-

arately. Regardless of the polymer type used for rein-

forcement, the exterior of the aerogel featured a

heterogeneous surface: in some parts, the surface is very

closed and the secondary particles are not individualized,

while on others the pearl necklace structure is well defined.

As explained earlier, due to the more complete condensa-

tion, the outer surface of the gels is more closed than their

Fig. 7 SEM micrographs of the aerogel’s microstructure

J Sol-Gel Sci Technol

123

interior. EDS analysis to this sample (Fig. 8), on each of

these very distinct regions, revealed that compared to the

average expected elemental composition of the silica

backbone, the sites where no individual particles are

observed have an excess of oxygen and carbon compared

with silicon, and the sites where individual particles are

observed feature a similar elemental composition to the one

expected on the silica backbone. These results confirm the

presence of the polymer in the outer surface, acting as a

binder of secondary silica particles. This is in agreement

with results obtained by TGA and observations of a more

rigid outer layer in the compression tests.

One-pot-grafted aerogels show a very regular pearl

necklace structure; the secondary particles are almost

spherical and have a narrow size distribution. They are also

smaller than the ones from the other aerogels. This very

regular microstructure is caused by the porogen effect of

the organic moieties during gelation, which controls the

particle growth (less extensive condensation) and pore size.

This observation is valid for both polymers, suggesting that

the molecular weight of the added macromer and monomer

is not a key factor relatively to the particle growth. These

samples feature less defined necks and more connections

between particles, suggesting a binding effect derived from

polymer incorporation.

4 Conclusions

Flexible, hydrophobic and lightweight reinforced aerogels

for insulation purposes were prepared. The mixture of

TMOS, MTMS and VTMS precursors has shown to be

efficient in tailoring the desired flexibility and hydropho-

bicity onto the final organically modified silica monoliths.

The reinforcement, performed with acrylates, was incor-

porated with the time-consuming gel soaking and stream-

line one-pot methodologies, in order to allow their

comparison.

Fig. 8 EDS analysis of two distinct areas on the outer surface of A_PP_S_R1.6 sample

J Sol-Gel Sci Technol

123

It has been demonstrated that, in spite of using a

monomer or a macromer in the synthesis, both generated

similar results, being the major differences between aero-

gels grafted using different methodologies. We found that

gel-soaking-grafted aerogels did not globally present rele-

vant improvements on their properties, being also highly

heterogeneous. The effects of monomer/macromer diffu-

sion through the silica backbone were significant in both

cases, despite the molecular weight differences. On the

contrary, one-pot aerogels showed improvements in almost

all evaluated properties, resulting in better heat insulation,

lighter, hydrophobic, homogeneous and flexible materials.

These improvements are in part due to the effect of the

presence of monomers during gelation. So, we conclude

that one-pot synthesis is a reliable method for obtaining

homogeneous monoliths suitable for insulation purposes.

Grafted aerogels became significantly more flexible as

Young’s modulus decreased by about 10 times when

compared to the non-reinforced aerogel. Comparing to

other polymer reinforced silica aerogels described in the

literature, the ones here studied have shown to be some of

the lightest, and with uncommonly low modulus for grafted

aerogels.

Acknowledgments This work was funded by FEDER funds throughthe Operational Programme for Competitiveness Factors—COM-

PETE and National Funds, through FCT—Foundation for Science and

Technology under the project PTDC/EQU–EPR/099998/2008—

GelSpace—Silica-Based Aerogels for Insulation of Spatial Devices.

References

1. Gurav JL, Jung I-K, Park H-H, Kang ES, Nadargi DY (2010)

Silica aerogel: synthesis and applications. J Nanomater

2010:1–11. doi:10.1155/2010/409310

2. Aegerter MA (2011) Aerogels handbook. In: Advances in sol-gel

derived materials and technologies. Springer. doi:10.1007/978-1-

4419-7589-8

3. Durães L, Ochoa M, Rocha N, Patrı́cio R, Duarte N, Redondo V,

Portugal A (2012) Effect of the drying conditions on the

microstructure of silica based xerogels and aerogels. J Nanosci

Nanotechnol 12(8):6828–6834. doi:10.1166/jnn.2012.4560

4. Soleimani Dorcheh A, Abbasi MH (2008) Silica aerogel; syn-

thesis, properties and characterization. J Mater Process Technol

199(1–3):10–26. doi:10.1016/j.jmatprotec.2007.10.060

5. Cai JY, Lucas S, Wang LJ, Cao Y (2011) Insulation properties of

the monolithic and flexible aerogels prepared at ambient pressure.

Adv Mater Res 391–392:116–120. doi:10.4028/www.scientific.

net/AMR.391-392.116

6. Ochoa M, Durães L, Beja A, Portugal A (2012) Study of the

suitability of silica based xerogels synthesized using

ethyltrimethoxysilane and/or methyltrimethoxysilane precursors

for aerospace applications. J Sol-Gel Sci Technol 61(1):151–160.

doi:10.1007/s10971-011-2604-7

7. Randall JP, Meador MA, Jana SC (2011) Tailoring mechanical

properties of aerogels for aerospace applications. ACS Appl

Mater Interfaces 3(3):613–626. doi:10.1021/am200007n

8. Maleki H, Durães L, Portugal A (2014) An overview on silica

aerogels synthesis and different mechanical reinforcing strate-

gies. J Non-Cryst Solids 385:55–74. doi:10.1016/j.jnoncrysol.

2013.10.017

9. Nguyen BN, Meador MA, Tousley ME, Shonkwiler B, McCorkle

L, Scheiman DA, Palczer A (2009) Tailoring elastic properties of

silica aerogels cross-linked with polystyrene. ACS Appl Mater

Interfaces 1(3):621–630. doi:10.1021/am8001617

10. Guo H, Nguyen BN, McCorkle LS, Shonkwiler B, Meador MAB

(2009) Elastic low density aerogels derived from bis[3-(tri-

ethoxysilyl)propyl]disulfide, tetramethylorthosilicate and

vinyltrimethoxysilane via a two-step process. J Mater Chem

19(47):9054. doi:10.1039/b916355g

11. Maleki H, Durães L, Portugal A (2014) Synthesis of lightweight

polymer-reinforced silica aerogels with improved mechanical and

thermal insulation properties for space applications. Microporous

Mesoporous Mater 197:116–129. doi:10.1016/j.micromeso.2014.

06.003

12. Mulik S, Sotiriou-Leventis C, Churu G, Lu H, Leventis N (2008)

Cross-linking 3D assemblies of nanoparticles into mechanically

strong aerogels by surface-initiated free-radical polymerization.

Chem Mater 20:5035–5046

13. Nguyen BC, Meador MA, Medoro A, Arendt V, Randall J,

McCorkle L, Shonkwiler B (2010) Elastic behavior of

methyltrimethoxysilane based aerogels reinforced with tri-iso-

cyanate. ACS Appl Mater Interfaces 2(5):1430–1443. doi:10.

1021/am100081a

14. Meador MA, Weber AS, Hindi A, Naumenko M, McCorkle L,

Quade D, Vivod SL, Gould GL, White S, Deshpande K (2009)

Structure-property relationships in porous 3D nanostructures:

epoxy-cross-linked silica aerogels produced using ethanol as the

solvent. ACS Appl Mater Interfaces 1(4):894–906. doi:10.1021/

am900014z

15. Meador MAB, Scherzer CM, Vivod SL, Quade D, Nguyen BN

(2010) Epoxy reinforced aerogels made using a streamlined

process. ACS Appl Mater Interfaces 2(7):2162–2168. doi:10.

1021/am100422x

16. Maleki H, Durães L, Portugal A (2015) Synthesis of mechani-

cally reinforced silica aerogels via surface-initiated reversible

addition-fragmentation chain transfer (RAFT) polymerization.

J Mater Chem A 3(4):1594–1600. doi:10.1039/c4ta05618c

17. Vareda J (2015) Smart polymer reinforced silica based aerogels.

MSc Thesis, Universidade de Coimbra

18. Durães L, Maia A, Portugal A (2015) Effect of additives on the

properties of silica based aerogels synthesized from

methyltrimethoxysilane (MTMS). J Supercrit Fluids 106:85–92.

doi:10.1016/j.supflu.2015.06.020

19. He Y-L, Xie T (2015) Advances of thermal conductivity models

of nanoscale silica aerogel insulation material. Appl Therm Eng

81:28–50. doi:10.1016/j.applthermaleng.2015.02.013

20. Durães L, Matias T, Patrı́cio R, Portugal A (2013) Silica based

aerogel-like materials obtained by quick microwave drying.

Mater Werkst 44(5):380–385. doi:10.1002/mawe.201300140

21. Al-Oweini R, El-Rassy H (2009) Synthesis and characterization

by FTIR spectroscopy of silica aerogels prepared using several

Si(OR)4 and R0 0Si(OR0)3 precursors. J Mol Struct919(1–3):140–145. doi:10.1016/j.molstruc.2008.08.025

J Sol-Gel Sci Technol

123

http://dx.doi.org/10.1155/2010/409310http://dx.doi.org/10.1007/978-1-4419-7589-8http://dx.doi.org/10.1007/978-1-4419-7589-8http://dx.doi.org/10.1166/jnn.2012.4560http://dx.doi.org/10.1016/j.jmatprotec.2007.10.060http://dx.doi.org/10.4028/www.scientific.net/AMR.391-392.116http://dx.doi.org/10.4028/www.scientific.net/AMR.391-392.116http://dx.doi.org/10.1007/s10971-011-2604-7http://dx.doi.org/10.1021/am200007nhttp://dx.doi.org/10.1016/j.jnoncrysol.2013.10.017http://dx.doi.org/10.1016/j.jnoncrysol.2013.10.017http://dx.doi.org/10.1021/am8001617http://dx.doi.org/10.1039/b916355ghttp://dx.doi.org/10.1016/j.micromeso.2014.06.003http://dx.doi.org/10.1016/j.micromeso.2014.06.003http://dx.doi.org/10.1021/am100081ahttp://dx.doi.org/10.1021/am100081ahttp://dx.doi.org/10.1021/am900014zhttp://dx.doi.org/10.1021/am900014zhttp://dx.doi.org/10.1021/am100422xhttp://dx.doi.org/10.1021/am100422xhttp://dx.doi.org/10.1039/c4ta05618chttp://dx.doi.org/10.1016/j.supflu.2015.06.020http://dx.doi.org/10.1016/j.applthermaleng.2015.02.013http://dx.doi.org/10.1002/mawe.201300140http://dx.doi.org/10.1016/j.molstruc.2008.08.025

Flexible acrylate-grafted silica aerogels for insulation purposes: comparison of reinforcement strategiesAbstractGraphical AbstractIntroductionExperimentalMaterialsSynthesis of the silica backbonePolymer reinforcement via gel soakingPreparation of reinforced gels via one-pot synthesisGels dryingAerogels characterization

Results and discussionKey propertiesChemical structureFTIR

Thermal characterizationMechanical behaviorMicrostructure

ConclusionsAcknowledgmentsReferences